Abstract: This paper presents an intelligent and user-friendly android-based application for the automatic detection of Alzheimer's disease and brain tumors. Neurological disorders like these pose significant diagnostic challenges due to their complexity and the critical need for early detection. Our project employs a hybrid approach, combining image processing and machine learning techniques, to provide rapid, preliminary diagnostic insights to both patients and medical professionals. The application features two primary modules: Brain Tumor Detection and Alzheimer's Prediction. The Brain Tumor module processes uploaded brain X-ray images, applying pre-processing steps like grayscale conversion, thresholding, and binarization. Feature extraction is then performed, followed by the use of a U-Net deep learning model to segment and classify the tumor, determining if the image indicates a normal state or a specific tumor type. The Alzheimer’s module operates on tabular health and lifestyle data, where users input structured features such as age, gender, ethnicity, education level, BMI, smoking and alcohol consumption habits, physical activity, and diet quality. These inputs undergo pre-processing through normalization and categorical encoding. The processed data is then analyzed using machine learning algorithms like Random Forest and Support Vector Machine to predict the likelihood of Alzheimer’s disease. By integrating both image-based and feature-based diagnostic methods into a single platform, this system significantly enhances the scope and accuracy of early detection. The dual-model architecture not only supports medical assessments but also empowers users with accessible health screening tools. Future enhancements are envisioned to include MRI integration, cloud deployment, and direct communication with healthcare providers. This intelligent application stands as a potential decision-support system in the domain of preventive neurological healthcare.

| DOI: 10.17148/IJIREEICE.2025.13510